Abstract:
Background: Diagnosis of tuberculosis (TB) in children is challenging due to non-specific
symptoms, difficulty of obtaining respiratory specimen and lack of sensitive diagnostic tests.
The available TB diagnostic tests require sputum specimen, which is very difficult to obtain
from young children. Other specimen like gastric aspirate (GA) can be applied to make
microbiological diagnosis. However, these methods are complex, relatively invasive and
often need the children to be hospitalized and overnight fasting. Thus, it needs alternative
non-invasive specimens for diagnosis of pulmonary TB in children unable to produce
sputum.
Objective: To evaluate the diagnostic performance of Xpert assay for diagnosis of TB from
stool specimen in children <15 years with presumptive pulmonary TB.
Method: A cross-sectional study was conducted at Jimma University Medical Center from 1 st
March to 30 th November 2019. Socio-demographic and clinical data of participants were
collected using structured questionnaires. Expectorated sputum or GA specimens were
collected and analyzed by Xpert assay and Lowenstein-Jensen ( LJ) culture. In addition, one
stool specimen was collected from each child and tested by Xpert assay. Based on clinical,
radiological and laboratory findings, patients were categorized into four groups: “confirmed
TB”, “probable TB”, “possible TB”, and “not TB”. Diagnostic performance of stool Xpert
was calculated with reference to LJ culture and composite reference standards (CRS).
Results: A total of 178 children were enrolled, 169 had complete microbiological and clinical
results. Male to female proportion was (81/88). The mean age was 3years, the majority 64.5%
of the participant age was <5 years old. The overall microbiologically and clinically
confirmed childhood TB proportion was 11.8% (20/169). Of these, 5.9% (10/169) had
confirmed TB, 2.4% (4/169) had probable TB, 3.6% (6/169) had possible TB and 88.2%
(149/169) had not-TB. Stool Xpert had sensitivity of 100% (95%CI: 66.4-100) and
specificity of 99.3% (95%CI: 96.2-100) compared to culture. Whereas, the overall sensitivity
and specificity of stool Xpert was 50% (95%CI: 27.2-72.8) and 100% (95%CI: 97.1-100)
compared to CRS.
Conclusions: Stool Xpert testing was improved sensitivity than GA Xpert testing that
can be easily implemented at lowest level of health care system. However, a negative Xpert
assay results may not exclude a diagnosis of TB in children and children with
strong clinical findings suggestive of TB should be started on ant-TB treatment
